from Hacker News

Ask HN: Are you switching to an AI related role?

by softirq on 12/30/23, 4:01 PM with 14 comments

With AI/AGI supposedly someday supplanting a subset of tech roles, are there folks who are retooling their careers to ride the wave rather than weather the storm? Are you taking courses, making internal/external job changes, adding AI sounding terms to your resumes? Or, maybe just abandoning simple coding work in favor of roles that are less replaceable via LLM autofill?

  • by ravenstine on 12/31/23, 1:37 PM

    No, because I don't think most companies even know what an AI role means yet. I'm not too keen on having my time wasted learning a bunch of already out of date crap for a company that can't muster a real plan around AI.

    Speaking of, I might have been more enthusiastic if I was a young man. In my 30s, I can hardly give a shit about spending my remaining years, before dedicating myself to a family, sitting in front of a screen trying to be relevant and learn some hackneyed Python scripts that implement already-outdated AI models. Fuck that shit. That's a young man's game. I already did that with JavaScript development, which now looks incredibly stable in contrast to AI.

    Most products integrating AI will be obsolete soon anyway. The idea that AI works better as an adjunct to existing applications seems reasonable in the short term, but LLMs will get more advanced to the point where doing things in a traditional way won't make sense. The people in charge of businesses right now are simply too set in their ways and are unwilling to actually understand this technology. They believe they will see riches by being a GPT-whisperer, and maybe they will in the short term.

  • by ajknzhol25 on 12/30/23, 6:04 PM

    I am the technical product owner for a product which is a deep learning based computer vision platform. It is a deep tech product featuring generative AI and foundation models designed to address our domain-specific challenges. I am located in the AI research division, working alongside research scientists holding PhDs and research engineers.

    My contributions are multi-faceted:

    1. Despite having a technical background in software engineering and lacking expertise in machine learning/deep learning, I am actively building my skills in DL. This enables me to understand the DL features (training, inference loops, segmenation, object detection and fine-tuning etc) required by the domain and how they should be developed in the backend and frontend.

    2. On occasions, I also delve directly into DL code to solve some minor bugs to enhance my understanding which allows me to make more informed decisions.

    The DL skill set proves beneficial during customer product onboarding sessions because the customers are technical and ask lots of deep learning questions. It also helps in the effective communication with both frontend and backend teams, helping them visualize the problem and determine the necessary workflows to support DL processes.

    So, yes. I am upskilling myself in DL to stay relevant in the AI product management.

  • by anonzzzies on 12/30/23, 9:34 PM

    I (and my company) are just incorporating LLMs into our products everywhere where they make sense. Which requires our engineers to know a lot more about them. We work with source code and data from clients and as such we have to use self hosted LLMs, so ops for ai related services is becoming a thing. I think this will become normal; it’s a tool to use next to other tools, for now.

  • by gtvwill on 12/30/23, 9:12 PM

    Im at a crunch point. I either leap further into IT and obviously leverage AI for work much as I do today. But I have no tickets and am a jack of all IT trades but a master of none.

    In IT the competition is high and the job security is low. So I either close my company and go work for a big dog and earn less than i would driving trucks, keep operating as a one man pty ltd and push to expand or bugger it all off and go building/laboring.

    Tbh bailing from IT as an industry is becoming more appealing. The industry sucks.

    Vendor lock in has gotten so bad folks don't implement good infrastructure rollout because of it.

  • by matrix87 on 12/30/23, 11:52 PM

    I dropped out of my swe job and replaced it with gig work, currently looking for a job as a prompt engineer. I think I'm going to go back to uni and get a degree in prompt engineering. The future is AI driven, whether people want to accept that or not

    Also my email is fuck_you@gmail.com if any journalists want to interview me, assuming journalists still exist in current year

  • by sk11001 on 12/30/23, 5:04 PM

    I'm an ML engineer and I don't agree with the assumption that AI-related roles are somehow different. I don't think they're necessarily in more demand, or pay more, or in lower risk of being automated away.

    I do think that, regardless of your job, AI assistance can be useful and worth getting used to, that's the biggest takeaway from the past 12 months.

  • by joshxyz on 12/31/23, 5:38 PM

    no, but im psitioning my projects in a way that can benefit rom those advancements.